1. No category

NeuroImage 91 (2014) 146–161
Contents lists available at ScienceDirect
NeuroImage
journal homepage: www.elsevier.com/locate/ynimg
Third order spectral analysis robust to mixing artifacts for mapping
cross-frequency interactions in EEG/MEG
F. Chella a,b,⁎, L. Marzetti a,b, V. Pizzella a,b, F. Zappasodi a,b, G. Nolte c
a
b
c
Department of Neuroscience and Imaging, “G. d'Annunzio” University of Chieti-Pescara, Chieti, Italy
Institute for Advanced Biomedical Technologies, “G. d'Annunzio” University Foundation, Chieti, Italy
Department of Neurophysiology and Pathophysiology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
a r t i c l e
i n f o
Article history:
Accepted 30 December 2013
Available online 10 January 2014
Keywords:
Functional connectivity
Cross-frequency interactions
Higher order spectral analysis
Bispectrum
Brain networks
Volume conduction
a b s t r a c t
We present a novel approach to the third order spectral analysis, commonly called bispectral analysis, of electroencephalographic (EEG) and magnetoencephalographic (MEG) data for studying cross-frequency functional
brain connectivity. The main obstacle in estimating functional connectivity from EEG and MEG measurements
lies in the signals being a largely unknown mixture of the activities of the underlying brain sources. This often
constitutes a severe confounder and heavily affects the detection of brain source interactions. To overcome this
problem, we previously developed metrics based on the properties of the imaginary part of coherency. Here,
we generalize these properties from the linear to the nonlinear case. Specifically, we propose a metric based
on an antisymmetric combination of cross-bispectra, which we demonstrate to be robust to mixing artifacts.
Moreover, our metric provides complex-valued quantities that give the opportunity to study phase relationships
between brain sources.
The effectiveness of the method is first demonstrated on simulated EEG data. The proposed approach shows a
reduced sensitivity to mixing artifacts when compared with a traditional bispectral metric. It also exhibits a better
performance in extracting phase relationships between sources than the imaginary part of the cross-spectrum for
delayed interactions. The method is then applied to real EEG data recorded during resting state. A crossfrequency interaction is observed between brain sources at 10 Hz and 20 Hz, i.e., for alpha and beta rhythms.
This interaction is then projected from signal to source level by using a fit-based procedure. This approach highlights a 10–20 Hz dominant interaction localized in an occipito-parieto-central network.
© 2014 Elsevier Inc. All rights reserved.
Introduction
Electroencephalography (EEG) and magnetoencephalography (MEG)
are noninvasive techniques which provide the opportunity to directly
measure ongoing brain activity with very high temporal but relatively
low spatial resolution. While in the past decades the main focus of
EEG/MEG studies was on the analysis of event related potentials, i.e. the
average brain response to a given stimulus, more recently the variability
of brain activity has attracted many researchers. The recent interest in
this field reflects the understanding that a mere localization of specific
brain activities is far from sufficient to understand how the brain
operates, but that it is necessary to study the brain as a network. In this
framework, the analysis of brain rhythms has been recognized as a promising approach since coherent neuronal activity has been hypothesized to
serve as a mechanism for neuronal communication (Fries, 2009; Gross
et al., 2006; Miller et al., 2009; Tallon-Baudry et al., 1996; Womelsdorf
and Fries, 2006).
⁎ Corresponding author at: Department of Neuroscience and Imaging, “G. d'Annunzio”
University of Chieti-Pescara, via dei Vestini 31, 66013 Chieti, Italy. Fax:+ 39 0871 3556930.
E-mail address: f.chella@unich.it (F. Chella).
1053-8119/$ – see front matter © 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.neuroimage.2013.12.064
The study of brain connectivity using noninvasive electrophysiological measurements like EEG or MEG also presents some problems which
still need to be faced. Most notably, the fact that the data are a largely
unknown mixture of the activities of the actual brain sources constitutes
a severe confounder. For instance, two sensors can record from the
same neural populations, opening the possibility for spurious interactions between sensors in the absence of true brain interactions. Though
the problem of mixing artifacts is well known (Nunez et al., 1997), it is
increasingly acknowledged and studied not only for channel data (often
referred to as volume conduction or field spread) (Srinivasan et al.,
2007; Winter et al., 2007) but also at the source level, i.e., after source
activities have been estimated from channel data using an inverse calculation (Schoffelen and Gross, 2009). Indeed, almost all the linear and
nonlinear methods used to analyze multivariate data for neuroscientific
applications (an excellent overview can be found in Pereda et al., 2005)
are highly sensitive to mixing artifacts.
To overcome the problem of volume conduction it was suggested to
exploit the fact that the propagation of electromagnetic fields is much
faster than neural communication: while phase shifts between electric
scalp potentials (EEG) or neuromagnetic fields (MEG) and the underlying source activity are too small to be observable (Stinstra and Peters,
1998), the temporal resolution of the data is still sufficient to capture
F. Chella et al. / NeuroImage 91 (2014) 146–161
phase shifts of neuronal signal propagation. This observation has been
exploited in the imaginary part of coherency (ImCoh), a measure of
brain connectivity which cannot be caused by mixtures of independent
sources (Nolte et al., 2004). However, in the presence of interacting
sources, the actual value still depends on how sources are mapped
into sensors (Nolte et al., 2004) or on source space (Sekihara et al.,
2011). This drawback was addressed for two sources using pairwise
measures with the lagged phase coherence (Pascual-Marqui, 2007b;
Pascual-Marqui et al., 2011) or with the weighted phase lag index
(Vinck et al., 2011).
Nonlinear methods to estimate correlations between power
addressing the issue of artifacts of volume conduction have also been
recently suggested. In Brookes et al. (2012), the authors address the
problem of field spread which generates spurious source space connectivity results. Using a seed based approach, the linear projection of the
seed voxel is first regressed out from the signals at the test voxel and,
then, power correlations are assessed both within and across multiple
frequency bands. Similarly, in Hipp et al. (2012), sensor signals are
orthogonalized before computing power envelope correlations at the
same or different frequencies, thus removing signal components that
share the same phase.
A noteworthy approach to the problem of volume conduction is
proposed in Gómez-Herrero et al. (2008). Here, after an initial principal
component analysis (PCA), the authors propose to subtract a linear
multivariate autoregressive model from sensor data to suppress all
time-delayed correlations, with the idea that all neural interactions
require a minimum delay. An independent component analysis (ICA)
is then applied to the residuals and the ICA mixing matrix is used to
model the effects of volume conduction (see also Hyvärinen et al.,
2010). This approach takes note of the fact that a direct application
of ICA to the data would be a conceptual contradiction to the objective of the research, namely studying causality relationships between
sources.
In this paper, we address the problem of mixing artifacts in relation
to the use of nonlinear methods for studying cross-frequency phasesynchronization between neuronal populations. Specifically we refer to
bispectral measures, which were developed and applied on EEG/MEG in
abundance (Darvas et al., 2009a, 2009b; Dumermuth et al., 1971;
Helbig et al., 2006; Jirsa and Müller, 2013; Schwilden, 2006; Wang et al.,
2007), and we examine the question of what information can be derived
from such measures that estimate true functional connectivity between
brain regions as opposed to mixing artifacts. Our new contribution is, essentially, the generalization to nonlinear methods of the concepts based
on the imaginary part of coherency to solve the problem of volume
conduction (Marzetti et al., 2008; Nolte et al., 2004, 2008, 2009). As will
be shown below, for linear measures (e.g., cross-spectra), the imaginary
part equals the antisymmetric part (apart from a factor ι, i.e., the imaginary unit) and the antisymmetry property is the more general principle
from which measures robust to artifacts of volume conduction can be derived for second-order (linear) and for third-order (nonlinear) moments.
In this way, antisymmetrized cross-bispectra can be used along with the
imaginary part of cross-spectra for identifying phase-locked brain areas
without being confounded by mixing artifacts, but with the important
difference that the former reflects the presence of brain rhythms locked
together at different frequencies, while the latter focuses on interactions
at the same frequency. Moreover, the proposed approach has also the advantage of improving, for a certain class of interactions, a limitation of the
imaginary part of the cross-spectrum, which cannot provide information
about relative phases, i.e. the phase difference of the activities of two
brain sources, in a way which is robust to artifacts of volume conduction.
Indeed, the imaginary part of the cross-spectrum is itself a real and not a
complex valued quantity, and real values do not contain information
about relative phases. Hence, the dilemma of linear measures is the fact
that possibly interesting quantities cannot be estimated in a way which
is robust to artifacts of volume conduction. On the contrary, the antisymmetric part of third order moments (cross-bispectra) is itself complex and
147
hence contains phase information which is not corrupted by noninteracting sources.
The paper is organized as follows. In the Material and methods section, we present the theory for cross-bispectral measures robust to
mixing artifacts. Specifically, we first recall the basic principles of the
imaginary part of cross-spectrum and, then, we introduce the antisymmetric part of the cross-bispectrum, discussing its properties with
regards to mixing artifacts. We describe a strategy to project the interaction from channel to source level by using a fit-based procedure. We
also discuss some examples of interpretation of the phase of crossbispectral measures. In the Result section, we first analyze the performance of our method in a simulation study, where we apply it on simulated EEG data. We then describe an example of an application of the
method to real EEG data. Finally, the Discussion section provides
remarks on the method features and on its ability to give an insight on
cross-frequency functional connectivity.
Material and methods
Theory for cross-bispectral measures robust to mixing artifacts
Cross-spectra and mixing artifacts
We first recall some principles of second order statistical analysis in
the frequency domain. The respective statistical moments, the elements
of the cross-spectral matrix S, are defined as
D
E
Sij ð f Þ ¼ X i ð f ÞX j ð f Þ
ð1Þ
where Xi(f) and Xj(f) are the Fourier coefficients of (eventually windowed) segments of data in channel i and channel j at frequency f, * denotes complex conjugation, and 〈 ⋅ 〉 denotes taking the expectation
value, i.e. taking the hypothetical average over an infinite number of
segments. Of course, the expectation value is unknown and will in general be estimated by a finite average over segments. Since S = S†, where
(∙)† denotes transpose and complex conjugation, S is an hermitian matrix. Complex coherency, C, is defined as the cross-spectrum normalized
by power, i.e. the diagonal elements of it:
Sij ð f Þ
C ij ð f Þ ¼ 1=2
:
Sii ð f ÞSjj ð f Þ
ð2Þ
It was argued that the imaginary part of the coherency is a useful
quantity to study brain interaction because it cannot be generated
from a superposition of independent sources (Nolte et al., 2004). For
later use we rederive this result assuming that the data have zero
mean which, if not vanishing, has to be subtracted from the raw data.
We now assume that all sources sk(f) are mapped instantaneously into
channels as
Xið f Þ ¼
X
aik sk ð f Þ
ð3Þ
k
with aik being real valued coefficients corresponding to the forward
mapping of the kth source to the ith channel. Then the cross-spectrum
can be written as
Sij ð f Þ ¼
X
k
2
aik ajk 〈jsk ð f Þj i þ
X
aik a jk′ sk ð f Þsk′ ð f Þ :
ð4Þ
′
k≠k
If we now assume that all sources are independent, the second term
on the right hand side in the above equation vanishes because for k ≠ k′
sk ð f Þsk′ ð f Þ ¼ hsk ð f Þi sk′ ð f Þ ¼ 0:
ð5Þ
Since the first term in Eq. (4) is real valued, a non-vanishing imaginary part of S must arise from interacting sources and can be used to
148
F. Chella et al. / NeuroImage 91 (2014) 146–161
study brain interactions. This applies equally to coherency where the
cross-spectrum was normalized with the real valued power. Note, that
the first term in Eq. (4) is not only real valued but also symmetric
with respect to switching the channel indices i and j.
The relation between the antisymmetric part of S, Sasym
= (Sij − Sji)/2
ij
and the imaginary part of S follows immediately from the fact that S is
hermitian, resulting in
n
o
asym
Sij ð f Þ ¼ ıIm Sij ð f Þ
ð6Þ
where ι denotes the imaginary unit. The above relation, Eq. (6), does not
exist for higher order moments, and, as it will be seen below, it is the symmetry property which can be generalized to higher order statistical
moments.
Bispectral analysis
Let Xi(f), Xj(f) and Xk(f) be the Fourier transforms of timeseries
recorded by three EEG (or MEG) channels i, j, and k. The crossbispectrum, which is essentially a measure of the quadratic phase
coupling between signal components at frequencies f1 , f 2, and
f1 + f2, is given by
D
E
Bijk ð f 1 ; f 2 Þ ¼ X i ð f 1 ÞX j ð f 2 ÞX k ð f 1 þ f 2 Þ
ð7Þ
where 〈 ⋅ 〉 represents the expectation value over a sufficiently large
number of signal segments.
The cross-bispectrum and other bispectral measures, e.g., bicoherence
(Nikias and Petropulu, 1993), have been widely used to detect and quantify the non-linear and non-Gaussian properties of EEG/MEG signals
(Dumermuth et al., 1971; Helbig et al., 2006; Schwilden, 2006; Wang
et al., 2007), with a particular interest in sleep studies (Barnett et al.,
1971; Bullock et al., 1997) and anesthesia monitoring (Johansen and
Sebel, 2000). Methods for nonlinear connectivity estimation based
on bispectral analysis of EEG recordings have also been developed
(e.g., Shils et al., 1996; Villa and Tetko, 2010). Almost all these methods,
however, are highly sensitive to mixing artifacts, and spurious interactions might be detected when analyzing data. A solution is to first localize the sources of brain activity and, then, apply the bispectral analysis
to source timecourses (see for example Darvas et al., 2009b). However,
the obtained results is in general only free of mixing artifacts if the
inverse solution accurately separates all sources which is hardly ever
the case.
Here, we introduce a new bispectral measure for sensor-level
connectivity estimation which is robust to mixing artifacts. We do not
replace the goal of a source-level analysis, since the interaction can
further be projected onto brain space, but, in such a way, the estimated
interaction is not affected by the validity of the inverse method. In particular, we look at the part of the cross-bispectrum which is antisymmetric to the permutation of two channel indices, e.g., i and k in the
following example1:
B½ij jjk ð f 1 ; f 2 Þ ¼ Bijk ð f 1 ; f 2 Þ−Bkji ð f 1 ; f 2 Þ:
ð8Þ
To see that the antisymmetrized cross-bispectrum of Eq. (8) cannot
be generated from independent sources, we assume that all sources,
sm(f), are independent and we insert that into Eq. (7)
X
aim ajm akm sm ð f 1 Þsm ð f 2 Þsm ð f 1 þ f 2 Þ þ coupling terms:
Bijk ð f 1 ; f 2 Þ ¼
m
ð9Þ
1
It is common to denote the antisymmetrizing operation on tensor indices by a bracket
notation in which [·] indicates antisymmetrization over a subset of indices included in the
brackets, e.g., Bi[jk] = Bijk − Bikj. In the event that indices to be antisymmetrized are not
adjacent to each other, the preceding notation is extended by using vertical lines to exclude indices from the antisymmetrization, i.e.: B[i|j|k] = Bijk − Bkji.
The ‘coupling terms’ contain expressions of the form 〈sm(f1)sn(f2)
s∗p(f1 + f2)〉 where not all indices m, n, p are identical, i.e. at least one
of the indices is different from the other two. If, e.g., this index is the
first one, m, and all sources are independent, this term vanishes
D
E
D
E
sm ð f 1 Þsn ð f 2 Þsp ð f 1 þ f 2 Þ ¼ hsm ð f 1 Þi sn ð f 2 Þsp ð f 1 þ f 2 Þ ¼ 0
ð10Þ
and likewise for any other of the indices.
Hence, for independent sources we get
Bijk ð f 1 ; f 2 Þ ¼
X
aim ajm akm sm ð f 1 Þsm ð f 2 Þsm ð f 1 þ f 2 Þ
ð11Þ
m
which is totally symmetric with respect to the three channel indices.
From this it follows that an antisymmetric combination with respect
to any pair of indices must vanish for independent sources. Below we
will restrict ourselves mainly to the case f1 = f2, namely we will focus
on cross-frequency interactions involving one frequency, f1 = f2 = f,
and its double, f1 + f2 = 2f. In this case only one antisymmetric
combination is relevant, namely antisymmetrization of the first and
last index as given in Eq. (8). Since Bijk(f1, f2) = Bjik(f1, f2) the
other two possible antisymmetric combinations are either redundant,
B[i|j|k](f1, f2) = Bj[ik](f1, f2), or vanish, B[ij]k(f1, f2) = 0. Note that in general B[i|j|k](f1, f2) is complex and allows to reconstruct phases also from
quantities which are robust to mixing artifacts. This is in sharp contrast
to second order statistics: when restricting to the imaginary (or antisymmetric) part of the cross-spectrum, any information on phases is lost.
The above symmetry property required the vanishing of coupling
terms for independent sources. We emphasize that this only holds up
to the third order statistical moments. For three (source) indices and if
two indices are equal, the third index is either different from both and
the corresponding (vanishing) mean can be factored out, or all indices
are equal and the contribution to the cross-bispectrum is totally symmetric. For the fourth or higher order, the coupling terms in general
do not vanish even if the sources are independent. E.g. for m ≠ k, for
independent sources sm and sk, and for f2 = −f1 one gets
sm ð f 1 Þsm ð f 2 Þsk ð f 3 Þsk ð f 1 þ f 2 þ f 3 Þ
¼ hsm ð f 1 Þsm ð f 2 Þi〈sk ð f 3 Þsk ðf 1 þ f 2 þ f 3 Þi ¼
ð12Þ
¼ jsm ð f 1 Þj2 i〈jsk ð f 3 Þj2 iN0
for non-vanishing signal power.
Normalization
It has been shown (Nikias and Petropulu, 1993; Van Ness, 1966)
that, for a sufficiently large sample size, the estimation of the crossbispectrum by using conventional methods provides unbiased estimates with asymptotic variance:
n n
oo
n n
oo
var Re Bijk ð f 1 ; f 2 Þ
¼ var Im Bijk ð f 1 ; f 2 Þ
¼
h
i
∼ 1 þ δ f 1 f 2 Sii ð f 1 ÞSjj ð f 2 ÞSkk ð f 1 þ f 2 Þ
ð13Þ
where δf1f2 is the Kronecker delta and Sii(f) = 〈Xi(f)X∗i (f)〉 is the power
spectrum of the ith process and similarly for j and k. This means that,
in general, the variance of the bispectral estimate increases with signal
power and, if the averaging is insufficient, large peaks may appear in
the cross-bispectrum because it is highly variable and not because
there is a significant phase coupling at that frequency pair. A normalization is, then, required in order to avoid improper interpretations. Here,
we divide the cross-bispectrum by the standard deviation of its estimate, as follows:
n
o
n
o
Re Bijk ð f 1 ; f 2 Þ
Im Bijk ð f 1 ; f 2 Þ
n n
oo þ i
n n
oo
bijk ð f 1 ; f 2 Þ ¼
std Re Bijk ð f 1 ; f 2 Þ
std Im Bijk ð f 1 ; f 2 Þ
ð14Þ
F. Chella et al. / NeuroImage 91 (2014) 146–161
with the standard deviation being computed as the conventional standard error of the mean, i.e.,
stdfg ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vD
u 2E
u −hi2
t
P
ð15Þ
where P denotes the number of segments into which the signals were
divided. As an alternative, since it is known from Eq. (13) that the variances of the real and imaginary parts of the cross-bispectrum are
asymptotically equal, a pooled estimator of the standard deviation can
ffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
n n oo2 n n oo2 be calculated as
std Re Bijk
þ std Im Bijk
=2 and
used in the denominator of Eq. (14). The adopted normalization allows
to suppress statistically meaningless estimates, thus assessing the robustness of obtained results. Similarly, the normalization of the antisymmetric part of the cross-bispectrum is achieved by dividing both
the real and the imaginary part of B[i|j|k] by the standard deviation of
their estimates, as follows:
n
o
n
o
Re B½ij jjk ð f 1 ; f 2 Þ
Im B½ij jjk ð f 1 ; f 2 Þ
n n
oo þ i
n n
oo : ð16Þ
b½ij jjk ð f 1 ; f 2 Þ ¼
std Re B½ij jjk ð f 1 ; f 2 Þ
std Im B½ij jjk ð f 1 ; f 2 Þ
Fit of the interaction
Once a statistically significant antisymmetric part of the crossbispectrum has been detected and, thus, an interaction has been isolated at sensor level, the signal can be projected onto the brain space using
a specific interaction model. In the following, we describe a model
consisting of two arbitrarily distributed interacting sources.
Given an N-channel EEG (or MEG) recording, if we denote by
s 1 (f) and s 2 (f) the source activities in the frequency domain and
by ai and bi, i = 1…N, respectively, their real-valued topographies,
i.e. Xi(f) = ais1(f) + bis2(f), then, according to Eqs. (7) and (8), the
antisymmetric part of the cross-bispectrum for each triplet of
channel recordings is given by:
h
i
B½ij jjk ð f 1 ; f 2 Þ ¼ ai a j bk −ak a j bi α ð f 1 ; f 2 Þ
h
i
þ ai b j bk −ak b j bi βð f 1 ; f 2 Þ
ð17Þ
where
α ð f 1 ; f 2 Þ ¼ s1 ð f 1 Þs1 ð f 2 Þs2 ð f 1 þ f 2 Þ − s2 ð f 1 Þs1 ð f 2 Þs1 ð f 1 þ f 2 Þ
ð18Þ
βð f 1 ; f 2 Þ ¼ s1 ð f 1 Þs2 ð f 2 Þs2 ð f 1 þ f 2 Þ − s2 ð f 1 Þs2 ð f 2 Þs1 ð f 1 þ f 2 Þ : ð19Þ
The unknown topographies are found by fitting the measured data
to the above model.In actual measurements, we usually select the frequency pair ef 1 ; ef 2 where the normalized antisymmetric part of the
cross-bispectra have their maximum. In such a way, only the more reliable bispectral estimates are considered for the model fit. Then, the fit
consists in finding 2N + 2 model parameters, namely 2N real-valued
parameters
for
and
the topographies
2 complex-valued parameters for
e ¼ β ef ; ef
e ¼ α ef 1 ; ef 2 and β
α
which best explain the available
1 2
set of N3 observations from each possible triplet of channels, under the
constraint of unit norm topographies. For simplicity, this problem
can be reformulated as an unconstrained optimization problem by setting
e as complex numbers of unit norm, which are representable by
e and β
α
real phase factors, and incorporating their magnitude in the topographies.
Hence, the problem reduces to the estimation of 2N + 2 real-valued
parameters.
The model fit can be achieved by means of standard nonlinear optimization techniques. Herein, we use a Levenberg–Marquardt algorithm
(Nocedal and Wright, 2006), minimizing the squared norm of the
149
difference between the model and the observations. Since the algorithm
is not guaranteed to converge to global minima, a standard multistart
approach is also used.
An important point is that the obtained solution is not unique. In
particular, we found that any linear combination of the estimated
topographies is also a valid solution for the fit, provided that the phase
factors are transformed accordingly (see Appendix A for details).
Hence, in general, the fit procedure does not provide the actual source
topographies, but an unknown superposition of them. Additional
assumptions are required to disentangle the actual sources. Here, we
apply the minimum overlap component analysis (MOCA), that is, we assume that the two sources are spatially separated (Marzetti et al., 2008).
This is a reasonable constraint, which allows to uniquely disentangle the
compound interacting system. To do this, the estimated topographies
are first localized by using a standard inverse solver, e.g., the eLORETA
reconstruction method (Pascual-Marqui, 2007a, 2009). According to
the above considerations, also the localized brain activities do not represent the actual sources, but an unknown superposition of them. Therefore, a constraint of minimum spatial overlap is applied to uniquely
disentangle the two sources. The actual topographies and the actual
phase factors are then retrieved accordingly.
Interpretation of cross-bispectral phases
Without considering confounders like artifacts of volume conduction
and additive noise, the interpretation of coherency is fairly simple: the
magnitude indicates the strength of the coupling and the phase indicates
a temporal relationship at a given frequency, even though the origin of
such a temporal relationship is in general unclear. The interpretation for
cross-bispectral values is much more complicated. Here, we want to
discuss in some more detail the interpretation of cross-bispectral phases.
Let us introduce a sketchy notation and denote by Φ the phase difference between the activity at two sites (which could be identical) at
two given frequencies (which could also be identical) from ranges
which are in a 1:2 ratio, e.g., in the alpha (f α, at about 10 Hz) or
beta (fβ, at about 20 Hz) range. Then Φ(fα1, fα2) denotes the phase
difference in the alpha range between site 1 and site 2, which could
be measured with coherency. Restricting to two frequencies, alpha
and beta, and to two sites, using coherency only two (non-vanishing)
phases can be calculated, Φ(fα1, fα2) and Φ(fβ1, fβ2), whereas, trivially,
Φ(fαk, fαk) = Φ(fβk, fβk) = 0 for any k.
For cross-bispectral measures six non-vanishing phases exist. Four of
them are fairly obvious: Φ(fα1, fβ1), Φ(fα2, fβ2), Φ(fα1, fβ2), and Φ(fα2, fβ1).
Two more arise from trivariate couplings expressed in the terms B122(f, f)
and B211(f, f) in Eq. (7), for f corresponding to the alpha frequency, and
can be denoted as Φ(fα1, fα2, fβ2) and Φ(fα2, fα1, fβ1) reflecting e.g. a
phase relationship between alpha in one site and a combination of
alpha and beta in the second site, namely Φ(fα1,fα2, fβ2) = ϕ(fα1) +
ϕ(fα2) − ϕ(fβ2) and Φ(fα2, fα1, fβ1) = ϕ(fα2) + ϕ(fα1) − ϕ(fβ1), with
ϕ(fα1) being the phase of the alpha activity at site 1 (and similarly for
others).
When restricting to antisymmetric parts, no phase relationships can
be detected for coherency because the imaginary parts are real valued.
This is different for cross-bispectral measures where also the antisymmetric parts are complex giving rise to two non-trivial phase relationships expressed in the terms B[1|1|2](f, f) and B[2|2|1](f, f) in Eq. (8).
These two phases are complicated to interpret, and we will here only
consider two special cases.
First, if the activities of the two sites have disjoint frequency content
and, say, the activity of site 1 is in the alpha range and the activity of site
2 is in the beta range, then B[2|2|1](f, f) = 0 and B[1|1|2](f, f) = B112(f, f). In
this case, the six phases reduce to one, and this one is also contained in
the antisymmetric part. We didn't observe that in real data, but we
emphasize that the corresponding interpretation requires that the
separation of sites in the frequency domain, to be discussed below, is
accurate.
150
F. Chella et al. / NeuroImage 91 (2014) 146–161
The second case is more complicated. We now assume that the
interaction between sites 1 and 2 consists in a time-delayed coupling,
i.e.,
′
s2 ðt Þ
¼C
′
s1 ðt−τ Þ⇒s2 ð f Þ
¼ C s1 ð f Þe
−ı2πfτ
ð20Þ
where s1′(t) and s2′(t) here denote the activities in the time domain of site
1 and site 2, s1(f) and s2(f) their respective Fourier transforms, C is a real
scale factor and τ is a non-zero time delay. In this case, not only the term
B[2|2|1](f, f) is non-vanishing for f corresponding to the alpha frequency,
but we also get that the ratio between B[1|1|2](f, f) and B[2|2|1](f, f) reflects
the phase difference in the alpha range, i.e., Φ(fα1, fα2). This result is
non-trivial and it is derived below for any two given frequencies f1
and f2. We first note, for later use, that the expressions for B[1|1|2] and
B[2|2|1] equate (apart from the sign) those for terms α and β in
Eqs. (18) and (19). Then, the following results will immediately
apply to the case of the estimation of the phase difference between
interacting brain sources, provided that they satisfy the assumption
made in Eq. (20). Indeed, by inserting Eq. (20) into Eqs. (18) and
(19), we get:
α ð f 1 ; f 2 Þ ¼ B½1j1j2 ð f 1 ; f 2 Þ ¼ C s1 ð f 1 Þs1 ð f 2 Þs1 ð f 1 þ f 2 Þ
h
i
ı2πð f 1 þ f 2 Þτ
−ı2π f 1 τ
−e
e
ð21Þ
2
βð f 1 ; f 2 Þ ¼ −B½2j2j1 ð f 1 ; f 2 Þ ¼ C s1 ð f 1 Þs1 ð f 2 Þs1 ð f 1 þ f 2 Þ
h
i
ı2π f τ
−ı2πð f 1 þ f 2 Þτ
e 1 −e
ð22Þ
from which we obtain
α 1 ι2π f 2 τ
¼ e
β C
ð23Þ
and, thus,
arg
!
B½1j1j2
α
eι2π f 2 τ
¼ arg −
¼ arg
≡ Δϕð f 2 ; τ Þ
β
B½2j2j1
ð24Þ
which is the phase difference between the two sites at frequency f2.
This is an important result since, if we compare Eq. (24) with the
estimation of Φ(fα1, fα2) by using coherency, we find that, whereas
the latter still depends on the real part of the cross-spectrum and,
thus, in actual measurements, is possibly affected by volume conduction effects, the former only requires the antisymmetric part of
the cross-bispectra to be estimated, which does not depend on
self-interactions. Hence, a more robust estimate is achieved. However, it is important to recall that Δϕ cannot be interpreted in terms of
causality because a discrete ambiguity remains due to the 2πperiodicity of the phase, that is, Δϕ(f2 , τ) = Δϕ(f 2 , τ ± n/f 2 ) for
n = 0,1,2,….
Of course, a nonlinear interaction between two or more sites could
be arbitrarily complicated, and the interpretation of phases depends
on the details of interaction. We only presented two special cases. A
general theory is beyond the scope of this paper.
current dipoles located on the cortical mantle. Each set of sources
included:
• interacting sources: 2 interacting dipoles exhibiting a cross-frequency
interaction between frequency components at f1 = f2 = 10 Hz and
f3 = f1 + f2 = 20 Hz. The interaction between the two sources was
simulated by first generating the timecourse of a self-interacting
source, say s1′(t), (namely whose frequency components at f1, f2 and
f3 are phase-synchronous) and then setting the timecourse of a second
source, say s2′(t), to a time-delayed copy of s1′(t), i.e., s2′(t) = s1′(t − τ).
The time delay τ was 10 ms for all simulation repetitions. In particular,
the timecourse of the self-interacting source, s1′(t), was generated by
summing the timecourse of a 10 Hz oscillator, obtained by bandpass filtering white Gaussian noise around 10 Hz, with 1 Hz bandwidth, and the timecourse of a 20 Hz oscillator which was phasesynchronous with the 10 Hz oscillator, i.e., obtained by squaring the
10 Hz oscillator followed by filtering around 20 Hz. For the filtering
we used a Butterworth filter, performing filtering in both the forward
and reverse directions to ensure zero phase distortion.
The choice of this particular interaction model was motivated by the
fact that the interaction between two sources having intrinsic nonlinear dynamics and being one a delayed copy of the other has both
linear and nonlinear features: the formers can be studied by using
cross-spectra and the latters by using cross-bispectra. This allowed a
comparison between the performances on phase delay estimation
obtained by using our method and linear methods based on the imaginary part of cross-spectrum.
• self-interacting sources: 4 uncorrelated self-interacting dipoles, 2 of
them exhibiting a self-interaction between frequency components at
f1 = f2 = 10 Hz and f3 = f1 + f2 = 20 Hz and the other 2 exhibiting
a self-interaction between frequency components at f1 = f2 = 3 Hz
and f3 = f1 + f2 = 6 Hz. These sources aimed to generate spurious
interaction results that we propose to remove by using the antisymmetric part of the cross-bispectrum rather than the traditional crossbispectrum.
• sources of background noise: 100 uncorrelated dipoles generating
white Gaussian noise.
Dipole locations and orientations were randomly chosen and varied
for different simulation repetitions. We also required a minimum
distance of 5 cm between dipoles (except for sources of background
noise).
This simulation study aimed at reproducing a 10–20 Hz interaction
which was also observed in real EEG data, corrupted by the signal
from additional self-interacting sources (both at the same and at different frequencies than interaction) and background noise. Two kinds of
signal-to-noise ratio were then defined: the signal-to-backgroundnoise ratio (SbNR) and the signal-to-self-interacting-noise-ratio
(SsiNR). Both SbNR and SsiNR were defined for the frequency component at 20 Hz, since it is always weaker than the frequency component
at 10 Hz. In particular, the SbNR (or SsiNR) at 20 Hz was calculated as
the ratio between the mean variance across channels of the signal generated by interacting sources and the mean variance of signal generated
by sources of background noise (or self-interacting sources). For each
simulation repetition, we analyzed all the combinations of these two
signal-to-noise ratios with both SbNR and SsiNR being 10, 1 and 0.1. A
total of 1000 simulation repetitions were performed.
Simulations
Generation of simulated data
A simulation study was carried out to quantitatively evaluate the
performance of the presented method. Ten minute EEG recordings,
sampled at 1 kHz, were generated for 118 channels located on the outermost layer of a standard realistic head model (Fonov et al., 2009,
2011). For each simulation repetition, a set of EEG sources was generated and channel recordings were numerically computed from sources by
solving the EEG forward problem. All sources were modeled as single
Bispectral analysis
Bispectral analysis was preceded by a data dimension reduction
using principal component analysis (PCA). This stage was necessary to
lower the computational costs required to estimate such a large number
of cross-bispectral estimates for each triplet of channel recordings and
for each frequency pair of a sampled 2D frequency domain. The original
set of 118 recordings was, thus, reduced to a smaller dataset containing
only the first 30 principal components. We also checked whether
dimension reduction did not appreciably attenuate the signal from
F. Chella et al. / NeuroImage 91 (2014) 146–161
interacting and self-interacting sources. The obtained signals were
divided into 1 second non-overlapping segments. Within each segment,
data were Fourier transformed, Hanning windowed and the crossbispectrum Bijk(f1, f2), its antisymmetric part B[i|j|k](f1, f2) and their normalized forms bijk(f1, f2) and b[i|j|k](f1, f2) were estimated for frequency
pairs (f1, f2) up to f1 + f2 = 30 Hz. The resulting frequency resolution
was 1 Hz on both the f1 and f2 axes.
Performance criteria
We first evaluated the performance of the antisymmetric part of the
cross-bispectrum in recognizing true from spurious interactions generated by self-interacting sources. To this purpose, we looked at differences between the spectral content of the traditional cross-bispectrum
Bijk(f1, f2) and its antisymmetric part B[i|j|k](f1, f2), e.g., to check whether
the latter was not sensitive to the self-interactions at 3–6 Hz.
The ability in recognizing truly interacting from self-interacting
sources
was also evaluated by analyzing separately the interaction at
ef ; ef
1 2 = (10 Hz, 10 Hz) namely at the frequency pair where the interaction was simulated, corresponding to the 10–20 Hz interaction.
In particular, we aimed to contrast the performance of the traditional
cross-bispectrum versus its antisymmetric part in extracting information about interacting sources, i.e., when a given interaction model is
fitted to experimental data. An interaction model for the antisymmetric
part of the cross-bispectrum consisting of two interacting brain sources
was given in Eq. (17). The corresponding model for the traditional
cross-bispectrum is given in the equation below, which has been
derived from Eq. (9) by expliciting the ‘coupling terms’:
Bijk ð f 1 ; f 2 Þ ¼
2 X
2 X
2
X
aim ajn akp ℬmnp ð f 1 ; f 2 Þ
ð25Þ
m¼1 n¼1 p¼1
where
D
E
ℬmnp ð f 1 ; f 2 Þ ¼ sm ð f 1 Þsn ð f 2 Þsp ð f 1 þ f 2 Þ ðfor
m ¼ 1; 2; n ¼ 1; 2; p ¼ 1; 2Þ:
ð26Þ
151
with S imag
12 ð f Þ transformed accordingly (see Marzetti et al., 2008, for a
more detailed description of this model). However, S imag
12 ð f Þ by itself
is not sufficient to provide information about phase difference between
sources, but the real part of cross-spectrum between sources, say
S real
12 ð f Þ, is also required. A second fit was then performed, aimed at
e
estimating S real
12 ð f Þ from the real part of cross-spectra at f = 10 Hz,
by using the following equation:
h
i
n
o
real
real
Re Sij ð f Þ ¼ ai1 a j1 S 11 ð f Þ þ ai1 a j2 þ ai2 a j1 S 12 ð f Þ
real
þ ai2 a j2 S 22 ð f Þ
ð28Þ
where
real
S mn ð f Þ ¼ Re sm ð f Þsn ð f Þ
ðfor m ¼ 1; 2; n ¼ 1; 2Þ :
ð29Þ
For the second fit, the source topographies were considered as
known parameters, having been previously estimated from quantities,
i.e., the imaginary parts of cross-spectra, which, differently from real
parts, were not corrupted by non-interacting sources.
We finally compared the estimates of phase differences obtained by
using the three methods, i.e.,
• Δϕ(10 Hz) = arg(β111/β121) for the traditional cross-bispectrum
• Δϕ(10 Hz) = arg(α/β) for the antisymmetric part of cross-bispectrum
real
• Δϕ(10 Hz) = −arctan(S imag
12 =S 12 ) for the cross-spectrum.
Application to real EEG data
The presented method was also applied to the analysis of real EEG
data recorded during resting state. The most pronounced rhythms that
occur at rest are the alpha (8–12 Hz) and the beta (16–30 Hz) oscillations and a cross-frequency phase synchronization between this two
components has been frequently observed (e.g. Nikulin and Brismar,
2006; Nikulin et al., 2011; Palva et al., 2005).
ð27Þ
Data acquisition and preprocessing
Twenty healthy subjects participated in the study (15 males; 5
females; age: 29.4 years; range: 21–39). Written consent and local
ethical committee agreement were obtained. Subjects were requested
to sit in a quite and dimly lit room and to fix a cross in front of them.
Measurements consisted of 10 min of continuous eyes-open resting
state activity. EEG signals were recorded by an EGI (Electrical Geodesics,
Inc.) 128-channel system, sampled at 1 kHz. The impedance of all the
electrodes was kept below 100 kΩ. The reference was set to Cz. Channel
locations on the scalp and the locations of 3 fiducial points (nasion, left
and right pre-auricular point) were measured by a 3D digitizer
(Polhemus, Colchester, VT, USA).
Data preprocessing was carried out before proceeding with the
bispectral analysis. Signals from electrodes located over the face and
neck were taken out because they were contaminated by muscular
activity, thus reducing the number of available channels to 98. Raw
data were band-pass filtered at 0.5–100 Hz. All recordings were visually
inspected and the segments of data containing spike artifacts were removed. An independent component analysis (ICA) was also performed
for instrumental and biological artifact removal. A particular attention
was paid to the rejection of artifacts from the eyes, heart and neck muscles. Finally, clean signals were re-referenced to infinity by the reference
electrode standardization technique (REST) (Yao, 2001) since this allows a more straightforward interpretation of scalp potentials and
bispectral maps in terms of underlying source interactions, similarly to
what has been demonstrated for EEG spectral imaging (Yao et al.,
2005) and coherency studies (Marzetti et al., 2007).
where S imag
12 ð f Þ ¼ Imfhs1 ð f Þs2 ð f Þig is a real coefficient being equal to the
imaginary part of cross-spectrum between source activities. The topographies estimated in this way also required to be demixed by MOCA,
Bispectral analysis
Bispectral analysis of EEG recordings was preceded by a PCA-based dimension reduction. The original set of 98 recordings was, thus, reduced to
Model equations were fitted to data at ef 1 ; ef 2 = (10 Hz, 10 Hz) and
the results were evaluated by comparing the correlation between the
true (known in simulation) and estimated topographies of interacting
sources. At this stage, i.e., before applying MOCA, the estimated topographies are still unique up to a two-dimensional mixing. Thus, the correlation was measured by the second of the two canonical correlation
coefficients between the two-dimensional spaces spanned by each pair
of topographies.
After applying MOCA (and the inverse method therein) to remove
the non-uniqueness of solution, the complex-valued model coefficients
in Eqs. (18), (19) and (26) contain information regarding the phase difference between interacting sources at 10 Hz. We then carried out a
comparison between the results on phase difference obtained by using
our method and those obtained by using: i) the traditional crossbispectrum, and ii) a linear method based on the estimation of the imaginary part of cross-spectrum. Point ii) required to re-analyze data by
using cross-spectra. To do this, we first detected the true linear interaction at 10 Hz by using only the imaginary parts of cross-spectra, since
these quantities are not biased by the contribution of non-interacting
sources (Nolte et al., 2004). Note that, for linear interactions, the role
of non-interacting sources is played by self-interacting sources in this
simulation. The imaginary part of cross-spectra at ef = 10 Hz was fitted
by the following two-source interaction model:
i
n
o h
imag
Im Sij ð f Þ ¼ ai1 a j2 −ai1 a j2 S 12 ð f Þ
152
F. Chella et al. / NeuroImage 91 (2014) 146–161
a smaller dataset containing only about the first 30 principal components.
Signals were then divided into 1 second non-overlapping segments containing continuous data. Within each segment, data were Fourier transformed, Hanning windowed and both the cross-bispectrum, Bijk(f1, f2),
and the antisymmetric part of the cross-bispectrum, B[i|j|k](f1, f2),were estimated for frequency pairs (f1, f2) up to f1 + f2 = 50 Hz. The resulting
frequency resolution was 1 Hz on both the f1 and f2 axes.
Moreover, as discussed above, it is important to normalize the
obtained estimates to reduce the variance effects and avoid detecting
spurious interactions. We then plotted the normalized bispectral estimates bijk(f1, f2) and b[i|j|k](f1, f2), defined in Eqs. (14) and (16), as a function of f1 and f2 and we considered those values whose magnitude was
larger than 5 as evidence of significant interaction. This threshold value
was chosen on the basis of a Montecarlo simulation testing the normalized bispectral values generated by pure i.i.d. Gaussian noise. Finally, the
frequency components showing a significant antisymmetric part of the
cross-bispectrum, and thus necessarily reflecting true brain interactions, were chosen and considered for further analysis.
Visualizing cross-bispectral maps
For each frequency pair, the whole bispectral data contain a large
amount of information, i.e., N × N × N estimates, where N is the number of recording channels (or principal components for reduced
datasets). Then, the visualization of the coupling between all channel
triplets might be quite complex. A simplified visualization is here proposed by looking at the coupling between channels pairs, rather than
triplets. In particular, we generate scalp maps of interaction by first
back-projecting the bispectral data from the principal component
space to the original sensor space and, then, by visualizing the connections between all pairs of electrodes by means of the procedure
described in Nolte et al. (2004) for coherency
mapping.
For a selected frequency pair ef 1 ; ef 2 , we then visualize only
the cross-bispectra (and their antisymmetric parts) which satisfy the
index equality i = j. For simplicity, we collect those data in a N × N
ef ; ef ),
e ¼ B ef ; ef (and B
e ¼B
matrix whose entries read: B
ik
iik
1
2
½ik
½ijijk
1
2
for all i and k. As an example, in Fig. 1 we show the magnitude of the
ef ; ef
(complex) cross-bispectra at frequency pair
= (10 Hz,
1
2
10 Hz) for one typical subject.
The single large circle is a 2D representation of the whole scalp. At
each electrode location, a small circle is placed representing the scalp
Fig. 1. Magnitude of the cross-bispectrum between all channel pairs, Biik, at frequency pair
(10 Hz, 10 Hz). Data from one subject.
and containing the magnitude of the cross-bispectrum of the respective
electrode (marked as a black dot) with all other electrodes. In other
e words, the ith small circle contains the ith row of the matrix B
ik . In
order to avoid overlaps, the small circles have been slightly shifted with
respect to their actual locations using a dedicated procedure (Nolte
et al., 2004).
Fit of the interaction
All the significant peaks of the antisymmetric part of the crossbispectrum reflect actual interactions occurring in thebrain.Among
these, only the most significant (i.e., at frequency pair ef 1 ; ef 2 corresponding to the largest value of b[i|j|k]) was chosen for localization. Then,
the two-source model in Eq. (17) was fitted to real data.
The localization of interacting sources was achieved by using a distributed cortically-constrained source model, without preferred orientation, and a standard realistic head model, obtained by averaging the MNI
of 152 normal adults (Fonov et al., 2009, 2011), for the EEG forward
problem solution. In particular, for each subject, the EEG electrodes
were co-registered to the standard head by aligning the fiducial points
and fitting each electrode location on the outermost shell of the threeshell head model. The EEG inverse problem was finally solved by using
the eLORETA reconstruction method (Pascual-Marqui, 2007a, 2009).
Results
Simulations
We first analyzed the differences between the spectral contents of
the traditional cross-bispectrum, Bijk(f1, f2), and of its antisymmetric
part, B[i|j|k](f1, f2), in order to evaluate the ability of the latter to recognize true from spurious interactions arising from self-interacting
sources. At this purpose, we looked at the normalized quantities
bijk(f1, f2) and b[i|j|k](f1, f2) since, as discussed previously, they reduce
the variance effects and avoid the detection of spurious interactions.
An example, chosen among all simulation repetitions, is shown in
Fig. 2. Here, the magnitude of bijk(f1, f2) (on the left) and b[i|j|k](f1, f2)
(on the right) are shown as a function of frequencies f1 and f2 up to
f1 + f2 = 30 Hz and for all combinations of SbNR and SsiNR. We recall
that N3 estimates are obtained for each frequency pair, where N is the
number of channels (i.e., principal components for reduced datasets).
To compactly represent this large amount of information, we decided
to show only the maximum of these N3 estimates in the plots.
We observe that, in all cases, the antisymmetric part of crossbispectrum suppresses the interaction peak shown by traditional crossbispectrum at frequency pair (3 Hz, 3 Hz), which was actually generated
only by self-interacting sources. Moreover, the magnitude of the antisymmetric part of the cross-bispectrum (whose value is indicated by color)
does not depend on SsiNR, contrary to the magnitude of the traditional
cross-bispectrum which increases for decreasing SsiNR, that is for increasing nonlinear self-interacting noise. This was taken as first evidence
that the our metric is not biased by the activity of self-interacting sources.
Conversely, both bijk(f1, f2) and b[i|j|k](f1, f2) are affected by background
noise level, as their magnitude decreases for decreasing SbNR.
The interaction detected by the two metrics at (10 Hz, 10 Hz) was
then fitted by model Eqs. (17) and (25). The results are summarized
in Fig. 3. In the top panels, the correlation between true and estimated
source topographies is shown for the traditional cross-bispectrum (on
the left) and for its antisymmetric part (on the right). In particular, we
plotted the histograms of 1 / (1 − r), with r being the second of the
two canonical correlation coefficients. Here, the correlation is used to
quantify the ability of the two metrics in identifying truly interacting
sources. We also recall that this value is computed before applying
MOCA (and the inverse method therein), while an example of inverse
reconstruction of interacting sources, as provided in output by MOCA,
is given in the bottom panels of Fig. 3.
F. Chella et al. / NeuroImage 91 (2014) 146–161
153
Fig. 2. Comparison between the spectral content of the normalized cross-bispectrum and the normalized antisymmetric part of the cross-bispectrum for all combinations of SbNR and
SsiNR. Data from one representative case among all simulation repetitions.
We first note that both correlation values and source reconstruction
are slightly affected by SbNR, while the main differences in using the traditional cross-bispectrum and its antisymmetric part depend on SsiNR.
For SsiNR = 10, namely for a low level of self-interacting noise, both
metrics successfully identify truly interacting sources. For SsiNR = 1,
namely for a medium level of self-interacting noise, the antisymmetric
part is still able to recognize truly interacting sources while the performances of the traditional cross-bispectrum become poorer, as revealed
by i) lower values of correlation coefficients and ii) the appearance of
ghosts in source reconstruction at the locations of self-interacting
sources. However, for SsiNR = 0.1, namely for a high level of selfinteracting noise, also the antisymmetric part of the cross-bispectrum is
biased by self-interacting sources. This was explained as evidence that,
in the presence of strong non-linear noise, the symmetric components
of bispectra may be not suppressed completely for finite length data.
We finally evaluated the ability of the antisymmetric part of the crossbispectrum to provide more robust estimates of the phase difference
between interacting sources than i) the traditional cross-bispectrum
and ii) the cross-spectrum. A comparison between the results obtained
by using the three methods are shown in Fig. 4. Here, data have been plotted in the range [0;1.2] and the vertical dashed lines mark the true simulated value, i.e., Δϕ = 0.628.
We note that, similarly to correlation coefficients, the results of
phase estimation are moderately affected by the background noise
level. In particular, we observe a slight shift of the main peak of the histograms towards lower values for a high background noise level
(SbNR = 0.1). Furthermore, the main differences between the three
methods rely only on the self-interacting noise level. In particular, for
a low self-interacting noise level (SsiNR = 10), all the three methods
provide reliable estimates of the phase difference. For an increasing
self-interacting noise level (SsiNR = 1) the performance of the antisymmetric part of the cross-bispectrum is slightly affected, while we
can observe: i) a higher variability of results for the traditional crossbispectrum, as measured by the standard deviation of the estimates;
and ii) a drift towards lower values for the main peak in the estimate
distribution of the cross-spectra, as was expected since the real parts
of cross-spectra are corrupted by self-interacting sources. Finally, for a
high self-interacting noise level (SsiNR = 0.1) both the traditional
cross-bispectrum and its antisymmetric part do not provide a stable
estimation of the phase difference, this result being dependent on the
performance of the fit.
Real EEG recordings
We first analyzed the antisymmetric part of the cross-bispectrum,
B[i|j|k](f1, f2), of all subjects in order to identify those who exhibited a
significant interaction signal. As outlined in the previous section, we
looked at its normalized form, b[i|j|k], (rather than B[i|j|k] itself) since,
while still proportional to the strength of the interaction, it suppresses
the spurious peaks resulting from high variance of the estimates. We
observed significant values of the antisymmetric part of the crossbispectrum in approximately one-half of all subjects (9 out of 20). For
these subjects, hereinafter referred to as Subjects 1–9, the antisymmetric part of the cross-bispectrum showed one significant peak at around
the frequency pair (f1, f2) = (10 Hz, 10 Hz), reflecting a crossfrequency interaction between sources of alpha- and beta-band signals,
namely at around f1 = f2 = 10 Hz and f3 = f1 + f2 = 20 Hz. However, a certain amount of intersubject variability in channel and source
topographies was also observed. The remaining subjects did not show
significant interactions and, thus, were dropped out from the analysis.
In the following, we will illustrate the results from a few subjects chosen
from those exhibiting a significant interaction signal and which best
reflect the observed intersubject variability.
From a comparison between the cross-bispectrum and its antisymmetric part, we found interesting differences which are described
below. By looking at their frequency profiles, we found that the antisymmetric part of the cross-bispectrum effectively suppressed peaks in the
cross-bispectrum, which were thus attributed to self-interaction phenomena. As an example, in Fig. 5 we show the results from Subjects 1
and 5, distinguished in two panels, for which the differences were more
clear. In each panel, both the normalized cross-bispectrum bijk(f1, f2)
(on the left) and the normalized antisymmetric part of the crossbispectrum b[i|j|k](f1, f2) (on the right) are shown. The plots read as
follows. In the upper part, the magnitudes of bijk(f1, f2) and b[i|j|k](f1, f2)
are shown as a function of frequencies f1 and f2 up to f1 + f2 = 50 Hz.
Moreover, since the main peaks always occurred at frequency pairs
located on the diagonal axis f 2 = f 1, the magnitudes of bijk (f 1, f2 )
and b[i|j|k](f1, f2) for f2 = f1 are shown separately in the bottom part
of each plot, where the reader can appreciate the values obtained
for each channel triplet.
In the top panel, we note that the normalized cross-bispectrum has a
broad maximum, peaking at two frequency pairs, i.e., (10 Hz, 10 Hz) and
(12 Hz, 12 Hz), which reflects a significant phase synchronization
154
F. Chella et al. / NeuroImage 91 (2014) 146–161
Fig. 3. Top panels: histograms of 1 / (1 − r), with r being the second of the two canonical correlation coefficients between true and estimated topographies of interacting sources, for all combinations of SbNR and SsiNR. Bottom panels: an example of
recovered brain sources (after MOCA and inverse method therein). The green dots mark the location of truly interacting sources; the green crosses mark the location of self-interacting sources.
F. Chella et al. / NeuroImage 91 (2014) 146–161
155
Fig. 4. Estimation of phase difference between interacting sources obtained by using: i) the traditional cross-bispectrum (top left), ii) the antisymmetric part of the cross-bispectrum (top right) and iii) cross spectra (bottom), for all combinations of
SbNR and SsiNR. The vertical dashed line denotes the true value of phase difference, i.e., Δϕ = 0.628. m and s denote the mean value and the standard deviation, respectively. Data from all simulation repetitions.
156
F. Chella et al. / NeuroImage 91 (2014) 146–161
Fig. 5. Comparison between the normalized cross-bispectrum bijk(f1, f2) (on the left) and the normalized antisymmetric part of the cross-bispectrum b[i|j|k](f1, f2) (on the right). Data from
Subject 1 (top panel) and Subject 5 (bottom panel).
between signal components lying in the alpha and beta bands. However, when looking at the normalized antisymmetric part of the
cross-bispectrum, we note that the interaction at (12 Hz, 12 Hz) is less
prominent, and thus the larger peak in the normalized cross-bispectrum
magnitude was possibly generated by self-interactions. Therefore, only
the frequency pair at around (10 Hz, 10 Hz) has to be considered for
connectivity analysis. Similarly, in the bottom panel, we observe that
the normalized antisymmetric part of the cross-bispectrum keeps only
the peak at around (11 Hz, 11 Hz), while it suppresses the peaks at
around (20 Hz, 20 Hz), (22 Hz, 11 Hz) and (11 Hz, 22 Hz), possibly generated by self-interacting processes, e.g., higher harmonic generation.
We further investigated the differences between the cross-bispectrum
and its antisymmetric part by analyzing their channel topographies. We
found that, as opposed to the cross-bispectrum, the spatial pattern of its
antisymmetric part was straightforwardly interpretable in terms of
underlying source interactions. An example is given in Fig. 6. Here,
the magnitude
of
the antisymmetric part of the cross-bispectrum,
i.e., B½ij jjk ef 1 ; ef 2 , at the frequency pair of maximum interaction,
F. Chella et al. / NeuroImage 91 (2014) 146–161
157
We finally carried out the localization of interacting sources by
using the strategy described in the Material and methods section. As
pointed out at the beginning of this section, a certain amount of individual variability was observed both at the channel and at the source
level. Then, to summarize the obtained results, in Fig. 8 we show the
localized electrical activities (top row) and the corresponding scalp
potentials (bottom row) for 5 subjects which best reflect the observed
variability.
In particular, Subject 1 showed an interacting system composed by
one source located in the occipital and central regions (source A) and
one source located in the parietal region (source B), thus confirming
what was expected from the analysis of the bispectral maps in Figs. 6
and 7. A similar correspondence between brain source localization and
bispectral maps was observed in all the other subjects.
Discussion
Fig. 6. Magnitude of the antisymmetric part of the cross-bispectrum between all channel
pairs, B[i|i|k], at frequency pair (10 Hz, 10 Hz). Data from Subject 1.
i.e., ef 1 ; ef 2 = (10 Hz, 10 Hz) is shown for Subject 1. We refer to
Fig. 1 for a comparison with the cross-bispectrum magnitude for
the same data set. We observe that both the cross-bispectrum
(Fig. 1) and its antisymmetric part (Fig. 6) reveal an interaction
between EEG activities in the occipital, parietal and central regions.
However, while in Fig. 1 we observe a high cross-bispectrum between
nearby electrodes, which is consistent with uniformly distributed independent sources, the antisymmetric part of the cross-bispectrum
(Fig. 6) reflects an interesting interaction structure: the parietal electrodes are mostly interacting with the occipital and the central ones,
and vice versa. This structure is more evident in the maps of Fig. 7,
where we display the antisymmetric part of the cross-bispectrum
(here, the real and the imaginary components are shown separately)
for a small subset of actual recording channels, i.e., at the approximate
locations of 20 electrodes of the international 10–20 system. In particular, we wish to point out that, within each small circle, the bispectral
map still has the spatial resolution of the 98 channel system. Moreover,
we marked the locations of all the 20 electrodes and highlighted with a
circled black dot the electrode with respect to which the crossbispectrum is calculated.
In summary, as opposed to the cross-bispectrum, the spatial pattern
of its antisymmetric part was able to reveal an interaction between
distinct sources, located in the parietal and occipital/central regions.
In this paper, we propose to use the antisymmetric part of third
order statistical moments, i.e. cross-bispectra, as a nonlinear measure
of functional connectivity. The reason comes from the observation that
these components cannot be generated from mixtures of independent
sources but solely reflect the existence of genuine interactions.
The proposed method allows to detect functional quadratic phase
coupling between brain sources and to localize their generators. A key
step to identify such quadratic phase couplings from cross-bispectra is
that of being able to reject spurious, i.e. non-significant, occurrences
which are often found in real data analysis. Indeed, the absence of a
quadratic phase coupling would produce no peaks in cross-bispectrum
only for data of infinite length. However, due to the finite length
encountered in practice, spurious peaks may appear in the crossbispectrum and in its antisymmetric part. To assess statistical significance of the peaks, we here adopted a normalization procedure in
which the cross-bispectral values are divided by the standard deviation
of their estimates. Specifically, the real and the imaginary part of complex cross-bispectral values are normalized separately. We observed
in simulations that a reasonable threshold value for the magnitude of
the normalized antisymmetric part of the cross-bispectrum to be distinguishable from values generated by pure Gaussian noise for finite data
length (e.g., 600 s) was found to be about 5. Other possible normalization could be used to the same end: e.g., the use of a pooled estimator
for the standard deviations of the real and the imaginary part of
cross-bispectra.
The above identified interactions were fitted to a two source model,
and their topographies were localized using a linear inverse method and
demixed using MOCA. This was based on the assumption that i) all observed interactions are pairwise; ii) the linear solver is adequate; and
Fig. 7. Real and imaginary components of the antisymmetric part of the cross-bispectrum, B[i|i|k], at frequency pair (10 Hz, 10 Hz) for a subset of actual recording channels. Data from
Subject 1.
158
F. Chella et al. / NeuroImage 91 (2014) 146–161
Fig. 8. Localization of interacting source activities (top row) and the corresponding scalp potentials (bottom row). The plot uses arbitrary color units. Data from Subjects 1–5.
iii) the sources have a minimum overlap (MOCA constraint). The last assumption is necessary because the topographies of the two sources for
each pair are only unique up to a mixing factor, mathematically similar
to the case of an eigenvector decomposition of a matrix with eigenvalues coming in pairs. To generalize our approach, all these three
assumptions could be replaced by others. First, we could extend the
interaction model to allow for more than two interacting sources. In
this case, the model equation to be fitted to data would include more
additive terms similar to those at the right-hand side of Eq. (17),
representing all possible relationships between brain sources taken in
pairs or triplets. Also in this case the non-uniqueness of the solution
needs to be faced. To this purpose, a generalization of the MOCAalgorithm able to take into account systems composed by more than
two sources is presented in Nolte et al. (2009). However, fitting this
more complicated model to the data might be less straightforward.
We did not investigate this possibility in the present work, but
multilinear decomposition methods (see, e.g., Kolda and Bader, 2009)
are a possible approach to face this issue. Second, for the inverse method, an alternative is to use RAP-MUSIC similarly to what is done by
Shahbazi Avarvand et al. (2012) to analyze the singular vectors of the
imaginary part of the cross-spectra. In contrast to the present approach,
RAP-MUSIC assumes dipolar sources, which, trivially, is an advantage if
the assumption is true, because the variance is reduced, and a disadvantage if not, because the bias is increased. In any event, also RAP-MUSIC
would fail in general if three sources are interacting with each other
but spanning only a two-dimensional subspace of the sensor space.
This situation is analogous to applying RAP-MUSIC to covariance matrices for perfectly correlated sources which would also lead to a collapse
of signal dimension in sensor space. With the present techniques such
cases cannot be treated. We will address these problems in a future
work.
The proposed method was first tested in simulations. The simulations showed that the antisymmetric part of the cross-bispectrum
allows to identify interacting sources in most circumstances where the
traditional cross-bispectrum is confounded by non-linear noisy sources.
We also observed that, if noisy sources are very strong, our metric is still
able to detect the presence of an interaction, but further analysis, such
as fitting an interaction model to the data, is possibly affected by the
noise bias. This is essentially due to the fact that, in the computation
of the antisymmetric part of cross-bispectra, the contribution of nonlinear non-interacting sources is suppressed in a statistical sense and,
if too large, it may be not completely removed. Simulations allowed us
also to compare phase estimations obtained with the antisymmetric
part of the cross-bispectrum to phase estimation from linear metrics
F. Chella et al. / NeuroImage 91 (2014) 146–161
that we previously introduced to study linear interactions robust to
mixing artifacts. This was achieved by simulating a particular interaction, namely the delayed coupling between two non-linear sources,
which implies the synchronization between rhythms both at the same
and at different frequencies and, thus, both linear and nonlinear measures can be used complementarily to analyze interaction dynamics.
The antisymmetric part of the cross-bispectrum offers an advantage
over the aforementioned linear approaches since it results in complexvalued quantities which contain phase information related to source
interaction. On the other hand, the physical interpretation of the phase
of the antisymmetric part of the cross-bispectrum might be complicated,
depending on the nature of the interaction. In this work, we gave explicit
interpretations in the following two cases: i) a pure nonlinear interaction
between two sources whose activities are contributed by oscillations at
different frequencies; ii) interaction between two sources with intrinsic
nonlinear dynamic and which are time-delayed copies. The latter case
was explored in our simulations showing that the antisymmetric part of
the cross-bispectrum exhibits better performances than the crossspectrum in recovering the phase difference between interacting sources.
The analysis pipeline was applied to real EEG data recorded during
eyes-open resting state. The approach revealed an interaction occurring
between brain sources at 10 Hz and 20 Hz in about half of the measured
subjects. A possible reason for this is given by a general practical problem of nonlinear methods: nonlinear phenomena are typically weaker
than linear ones. Indeed, the observed interactions were mainly occurring between 10 Hz and 20 Hz, which are the most pronounced
rhythms at rest. Even though couplings between other frequencies are
possibly observable, the spectral content is surely richer when using linear methods. Moreover, the observed topographies for interacting
sources are less stable than those usually obtained from second order
moments. In our data interacting sources are localized, for all subjects
showing meaningful results for the antisymmetric part of crossbispectra, in occipito-parietal and central areas, consistently with previous observations (Palva et al., 2005). Indeed, when dealing with nonlinear approaches, we are faced with a trade-off: if a nonlinear signal can
be observed, a deeper insight into brain dynamic is possible but such
signals are more difficult to detect consistently in all subjects. Nevertheless, nonlinear dynamics might represent a key mechanism for brain
interaction: cross-frequency phase synchronization have been recognized to be a mechanism for different networks to interact and integrate
spectrally distributed processing (Palva et al., 2005). In this framework,
cross-frequency phase synchronization between alpha and beta rhythms
is already known to occur at rest (Nikulin and Brismar, 2006; Nikulin
et al., 2011) and to be modulated by task and shown to be involved in
the coordination between attention and perception (Palva et al., 2005).
Although, a comprehensive comparison with other nonlinear
methods is beyond the scope of this work, we want to emphasize that
our method is based on third order moments which are a ‘pure’ nonlinearity in the sense that for Gaussian distributed data these moments
vanish apart from random fluctuations. This is a substantial difference
to many other nonlinear measures, e.g., the phase locking value
(Lachaux et al., 1999) which does not vanish for Gaussian distributions
and which measures properties of phase distributions well consistent
with Gaussian distributions (Aydore et al., 2013). It is hence unclear
whether an interaction observed with phase locking cannot equally
be observed with coherence. Similarly, transfer entropy (Schreiber,
2000), which is a nonlinear measure of effective connectivity, is formally equivalent to Granger Causality for Gaussian distributed data. Also,
correlations of the power of signals at the same frequency can always
be expressed as functions of complex coherency which follows from
the fact that cross-spectra form a complete and sufficient statistics for
Gaussian distributed and zero-mean stationary processes. Methods
based on correlations of power designed to remove artifacts of volume
conduction were suggested in Brookes et al. (2012) and in Hipp et al.
(2012). These methods are based on the orthogonalization of segments
of data between pairs of sensors in the time-domain and in the
159
frequency domain, respectively. It is an open question, under which
conditions this orthogonalization actually removes rather than just attenuates artifacts of volume conduction.
Functional connectivity based on the antisymmetric part of crossbispectra for EEG and MEG can also be estimated directly between
brain source activities. An appealing framework would be the characterization of resting state network (RSNs) (Deco and Corbetta, 2011) phenomenon. Frequency specific coupling in RSNs have been recently
studied with the imaginary part of coherency (Martino et al., 2011;
Marzetti et al., 2013) as well as with correlation of power (Betti et al.,
2013; Brookes et al., 2011; de Pasquale et al., 2010, 2012; Hipp et al.,
2012). All these approaches have investigated coupling between brain
regions at the same frequency, with an emphasis on phase coupling or
on amplitude coupling. Conversely, the proposed method, while being
intrinsically robust to mixing artifacts, is designed to reveal crossfrequency interactions or might serve as a basis for getting further
insight into the study of phase delayed coupling at a given frequency.
In conclusion, we believe that the proposed method might provide a
new framework for further understanding the role of brain rhythms in
the functional integration and segregation of brain networks.
Acknowledgments
This work was partially supported by the Italian Ministry of Education, University and Research (PRIN 2010–2011 n. 2010SH7H3F_006
‘Functional connectivity and neuroplasticity in physiological and pathological aging’) and by grants from the EU (ERC-2010-AdG-269716), the
DFG (SFB 936/A3), and the BMBF (031A130). The author L.M. was
partially funded by the ‘Mapping the Human Connectome Structure,
Function, and Heritability’ (1U54MH091657-01) from the 16 National
Institutes of Health Institutes and Centers that support the NIH Blueprint
for Neuroscience Research.
Conflict of interest
The authors disclose that there is no actual or potential conflict of
interest.
Appendix A
Let us consider the antisymmetric part of the cross-bispectrum
between three channel recordings, namely i, j and k, as defined in
Eq. (17):
h
i
B½ij jjk ð f 1 ; f 2 Þ ¼ ai a j bk −ak a j bi α ð f 1 ; f 2 Þ
h
i
þ ai b j bk −ak b j bi βð f 1 ; f 2 Þ
ðA:1Þ
where α(f1, f2) and β(f1, f2) are complex valued functions and ai and
bi, i = i, j, k, are the real valued topographies of the two interacting
sources at those channels. Let us now replace ai and bi, i = i, j, k, by a
linear mixture of them, i.e.,
(
′
ai ¼ Aai þ Bbi
′
bi ¼ Cci þ Dbi
ðA:2Þ
with arbitrary coefficients A, B, C and D. For simplicity, we will use
the matrix notation by collecting the coefficients in a 2 × 2 mixing matrix Γ, i.e.,
′
ai
′
bi
!
¼
A
C
B
D
ai
bi
a
¼Γ i :
bi
ðA:3Þ
160
F. Chella et al. / NeuroImage 91 (2014) 146–161
In general, the above substitution also changes the antisymmetric
part of the cross-bispectrum, which depends on source topographies.
Therefore, if we want to keep unchanged the value of B[i|j|k], we
also have to replace α(f1, f2) and β(f1, f2) by new complex functions
α′(f1,f2) and β′(f1,f2), i.e.,
h
i
′ ′ ′
′ ′ ′
′
B½ij jjk ð f 1 ; f 2 Þ ¼ ai a j bk −ak a j bi a ð f 1 ; f 2 Þ
h
i
′ ′ ′
′ ′ ′
′
þ ai b j bk −ak b j bi β ð f 1 ; f 2 Þ:
ðA:4Þ
In particular, by combining the Eqs. (A.2) and (A.4), we get:
h
i
′
′
B½ij jjk ð f 1 ; f 2 Þ ¼ ai ; a j ; bk ; −; ak ; a j ; bi Aα ð f 1 ; f 2 Þ þ Cβ ð f 1 ; f 2 Þ ðAD−BC Þ þ
h
i
′
′
ai ; b j ; bk ; −; ak ; b j ; bi Bα ð f 1 ; f 2 Þ þ Dβ ð f 1 ; f 2 Þ ðAD−BC Þ:
ðA:5Þ
Then, a straightforward comparison of Eqs. (A.5) and (A.1) yields to:
8
< α ð f 1 ; f 2 Þ ¼ Aα ′ ð f 1 ; f 2 Þ þ Cβ′ ð f 1 ; f 2 Þ ðAD−BC Þ
: βð f ; f Þ ¼ Bα ′ ð f ; f Þ þ Dβ′ ð f ; f Þ ðAD−BC Þ
1
2
1
2
1 2
ðA:6Þ
or, in matrix notation,
αð f 1 ; f 2 Þ
βð f 1 ; f 2 Þ
T
¼ detfΓgΓ
′
α ð f 1; f 2Þ
′
β ð f 1; f 2Þ
!
ðA:7Þ
where ΓT is the transpose of the mixing matrix.
Hence, according to Eq. (A.7), the antisymmetric part of the crossbispectrum is invariant to the arbitrary linear combination of source
topographies defined in Eq. (A.2), if we simultaneously take α′(f1,f2)
and β′(f1,f2) such that:
′
α ð f 1; f 2Þ
′
β ð f 1; f 2Þ
!
¼
ðΓÞ−1
detfΓg
αð f 1 ; f 2 Þ
:
βð f 1 ; f 2 Þ
ðA:8Þ
References
Aydore, S., Pantazis, D., Leahy, R.M., 2013. A note on the phase locking value and its properties. NeuroImage 74, 231–244 (URL http://www.sciencedirect.com/science/article/
pii/S1053811913001286).
Barnett, T.P., Johnson, L.C., Naitoh, P., Hicks, N., Nute, C., 1971. Bispectrum analysis of electroencephalogram signals during waking and sleeping. Science 172 (3981), 401–402
(URL http://www.sciencemag.org/content/172/3981/401.abstract).
Betti, V., Della Penna, S., de Pasquale, F., Mantini, D., Marzetti, L., Romani, G.L., Corbetta, M.,
2013. Natural scenes viewing alters the dynamics of functional connectivity in the
human brain. Neuron 79 (4), 782–797 (URL http://www.sciencedirect.com/science/
article/pii/S0896627313005382).
Brookes, M.J., Hale, J.R., Zumer, J.M., Stevenson, C.M., Francis, S.T., Barnes, G.R., Owen, J.P.,
Morris, P.G., Nagarajan, S.S., 2011. Measuring functional connectivity using MEG:
methodology and comparison with fcMRI. NeuroImage 56 (3), 1082–1104 (URL
http://www.sciencedirect.com/science/article/pii/S10538119110021).
Brookes, M.J., Woolrich, M.W., Barnes, G.R., 2012. Measuring functional connectivity in MEG:
a multivariate approach insensitive to linear source leakage. NeuroImage 63 (2),
910–920 (URL http://www.sciencedirect.com/science/article/pii/S1053811912003291).
Bullock, T.H., Achimowicz, J.Z., Duckrow, R.B., Spencer, S.S., Iragui-Madoz, V.J., 1997.
Bicoherence of intracranial EEG in sleep, wakefulness and seizures. Electroencephalogr.
Clin. Neurophysiol. 103 (6), 661–678 (URL http://www.sciencedirect.com/science/
article/pii/S00134694970008).
Darvas, F., Miller, K.J., Rao, R.P.N., Ojemann, J.G., 2009a. Nonlinear phase-phase crossfrequency coupling mediates communication between distant sites in human
neocortex. J. Neurosci. 29 (2), 426–435 (URL http://www.jneurosci.org/content/29/
2/426.abstract).
Darvas, F., Ojemann, J.G., Sorensen, L.B., 2009b. Bi-phase locking — a tool for probing nonlinear interaction in the human brain. NeuroImage 46 (1), 123–132 (URL http://
www.sciencedirect.com/science/article/pii/S1053811909000949).
de Pasquale, F., Della Penna, S., Snyder, A.Z., Lewis, C., Mantini, D., Marzetti, L., Belardinelli,
P., Ciancetta, L., Pizzella, V., Romani, G.L., Corbetta, M., 2010. Temporal dynamics of
spontaneous MEG activity in brain networks. Proc. Natl. Acad. Sci. 107 (13),
6040–6045 (URL http://www.pnas.org/content/107/13/6040.abstract).
de Pasquale, F., Della Penna, S., Snyder, A.Z., Marzetti, L., Pizzella, V., Romani, G.L., Corbetta,
M., 2012. A cortical core for dynamic integration of functional networks in the resting
human brain. Neuron 74 (4), 753–764 (URL http://www.sciencedirect.com/science/
article/pii/S08966273120034).
Deco, G., Corbetta, M., 2011. The dynamical balance of the brain at rest. Neuroscientist 17
(1), 107–123 (URL http://nro.sagepub.com/content/17/1/107.abstract).
Dumermuth, G., Huber, P.J., Kleiner, B., Gasser, T., 1971. Analysis of the interrelations
between frequency bands of the EEG by means of the bispectrum a preliminary
study. Electroencephalogr. Clin. Neurophysiol. 31 (2), 137–148 (URL http://www.
sciencedirect.com/science/article/pii/00134694719018).
Fonov, V.S., Evans, A.C., McKinstry, R.C., Almli, C.R., Collins, D.L., 2009. Unbiased
nonlinear average age-appropriate brain templates from birth to adulthood.
NeuroImage 47 (Suppl. 1), S102, (URL http://www.sciencedirect.com/science/
article/pii/S1053811909708845).
Fonov, V., Evans, A.C., Botteron, K., Almli, C.R., McKinstry, R.C., Collins, D.L., 2011. Unbiased
average age-appropriate atlases for pediatric studies. NeuroImage 54 (1), 313–327
(URL http://www.sciencedirect.com/science/article/pii/S1053811910010062).
Fries, P., 2009. Neuronal gamma-band synchronization as a fundamental process in cortical computation. Annu. Rev. Neurosci. 32 (1), 209–224 (pMID: 19400723.URL http://
www.annualreviews.org/doi/abs/10.1146/annurev.neuro.051508.135603).
Gómez-Herrero, G., Atienza, M., Egiazarian, K., Cantero, J.L., 2008. Measuring directional
coupling between EEG sources. NeuroImage 43 (3), 497–508 (URL http://www.
sciencedirect.com/science/article/pii/S1053811908008549).
Gross, J., Schmitz, F., Schnitzler, I., Kessler, K., Shapiro, K., Hommel, B., Schnitzler, A., 2006.
Anticipatory control of long-range phase synchronization. Eur. J. Neurosci. 24 (7),
2057–2060. http://dx.doi.org/10.1111/j.1460-9568.2006.05082.x.
Helbig, M., Schwab, K., Leistritz, L., Eiselt, M., Witte, H., 2006. Analysis of time-variant quadratic phase couplings in the tracé alternant EEG by recursive estimation of 3rd-order
time–frequency distributions. J. Neurosci. Methods 157 (1), 168–177 (URL http://
www.sciencedirect.com/science/article/pii/S01650270060018).
Hipp, J.F., Hawellek, D.J., Corbetta, M., Siegel, M., Engel, A.K., 2012. Large-scale cortical
correlation structure of spontaneous oscillatory activity. Nat. Neurosci. 15, 884–890.
http://dx.doi.org/10.1038/nn.3101.
Hyvärinen, A., Zhang, K., Shimizu, S., Hoyer, P.O., 2010. Estimation of a structural vector
autoregression model using non-Gaussianity. J. Mach. Learn. Res. 11, 1709–1731
(Aug., URL http://dl.acm.org/citation.cfm?id=1756006.1859907).
Jirsa, V., Müller, V., 2013. Cross-frequency coupling in real and virtual brain networks.
Front. Comput. Neurosci. 7 (78) (URL http://www.frontiersin.org/computational_
neuroscience/10.3389/fncom.2013.00078/abstract).
Johansen, J.W., Sebel, P.S., 2000. Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology 93, 1336–1344 (URL http://
journals.lww.com/anesthesiology/Fulltext/2000/11000/Development_and_Clinical_
Application_of.29.aspx).
Kolda, T., Bader, B., 2009. Tensor decompositions and applications. SIAM Rev. 51 (3),
455–500. http://dx.doi.org/10.1137/07070111X.
Lachaux, J.-P., Rodriguez, E., Martinerie, J., Varela, F.J., 1999. Measuring phase synchrony in
brain signals. Hum. Brain Mapp. 8 (4), 194–208. http://dx.doi.org/10.1002/(SICI)
1097-0193(1999) 8:4b194::AID-HBM4N3.0.CO;2-C.
Martino, J., Honma, S.M., Findlay, A.M., Guggisberg, A.G., Owen, J.P., Kirsch, H.E., Berger, M.S.,
Nagarajan, S.S., 2011. Resting functional connectivity in patients with brain tumors in
eloquent areas. Ann. Neurol. 69 (3), 521–532. http://dx.doi.org/10.1002/ana.22167.
Marzetti, L., Nolte, G., Perrucci, M.G., Romani, G.L., Del Gratta, C., 2007. The use of standardized infinity reference in EEG coherency studies. NeuroImage 36 (1), 48–63
(URL http://www.sciencedirect.com/science/article/pii/S105381190700122X).
Marzetti, L., Del Gratta, C., Nolte, G., 2008. Understanding brain connectivity from EEG
data by identifying systems composed of interacting sources. NeuroImage 42 (1),
87–98 (URL http://www.sciencedirect.com/science/article/pii/S1053811908005971).
Marzetti, L., Della Penna, S., Snyder, A., Pizzella, V., Nolte, G., de Pasquale, F., Romani, G.L.,
Corbetta, M., 2013. Frequency specific interactions of MEG resting state activity within and across brain networks as revealed by the multivariate interaction measure.
NeuroImage 79, 172–183 (URL http://www.sciencedirect.com/science/article/pii/
S10538119130040).
Miller, K.J., Weaver, K.E., Ojemann, J.G., 2009. Direct electrophysiological measurement of
human default network areas. Proc. Natl. Acad. Sci. 106 (29), 12174–12177 (URL
http://www.pnas.org/content/106/29/12174.abstract).
Nikias, C.L., Petropulu, A.P., 1993. Higher-order spectra analysis: a nonlinear signal
processing framework. Prentice Hall Signal Processing Series.PTR Prentice Hall (URL
http://books.google.it/books?id=D_hSAAAAMAAJ).
Nikulin, V.V., Brismar, T., 2006. Phase synchronization between alpha and beta oscillations
in the human electroencephalogram. Neuroscience 137 (2), 647–657 (URL http://
www.sciencedirect.com/science/article/pii/S0306452205011711).
Nikulin, V.V., Nolte, G., Curio, G., 2011. A novel method for reliable and fast extraction of
neuronal EEG/MEG oscillations on the basis of spatio-spectral decomposition.
NeuroImage 55 (4), 1528–1535 (URL http://www.sciencedirect.com/science/article/
pii/S1053811911000930).
Nocedal, J., Wright, S., 2006. Numerical optimization. Springer Series in Operations
Research and Financial Engineering.Springer (URL http://books.google.it/books?
id=eNlPAAAAMAAJ).
Nolte, G., Bai, O., Wheaton, L., Mari, Z., Vorbach, S., Hallett, M., 2004. Identifying true brain
interaction from EEG data using the imaginary part of coherency. Clin. Neurophysiol.
115 (10), 2292–2307 (URL http://www.sciencedirect.com/science/article/pii/
S13882457040019).
Nolte, G., Ziehe, A., Nikulin, V.V., Schlögl, A., Krämer, N., Brismar, T., Müller, K.-R., 2008.
Robustly estimating the flow direction of information in complex physical systems.
Phys. Rev. Lett. 100, 234101. http://dx.doi.org/10.1103/PhysRevLett.100.234101 (Jun).
Nolte, G., Marzetti, L., Valdes Sosa, P., 2009. Minimum overlap component analysis
(MOCA) of EEG/MEG data for more than two sources. J. Neurosci. Methods 183 (1),
72–76 (URL http://www.sciencedirect.com/science/article/pii/S0165027009003690).
F. Chella et al. / NeuroImage 91 (2014) 146–161
Nunez, P.L., Srinivasan, R., Westdorp, A.F., Wijesinghe, R.S., Tucker, D.M., Silberstein, R.B.,
Cadusch, P.J., 1997. EEG coherency: I: statistics, reference electrode, volume conduction,
laplacians, cortical imaging, and interpretation at multiple scales. Electroencephalogr.
Clin. Neurophysiol. 103 (5), 499–515 (URL http://www.sciencedirect.com/science/
article/pii/S00134694970006).
Palva, J.M., Palva, S., Kaila, K., 2005. Phase synchrony among neuronal oscillations in the
human cortex. J. Neurosci. 25 (15), 3962–3972 (URL http://www.jneurosci.org/
content/25/15/3962.abstract).
Pascual-Marqui, R.D., 2007a. Discrete, 3D distributed, linear imaging methods of electric
neuronal activity. Part 1: exact, zero error localization (Oct., ArXiv e-prints).
Pascual-Marqui, R.D., 2007b. Instantaneous and lagged measurements of linear and nonlinear dependence between groups of multivariate time series: frequency decomposition (Nov., ArXiv e-prints).
Pascual-Marqui, R.D., 2009. Quantitative EEG analysis: methods and clinical applications.
In: Tong, S., Thakor, N.V. (Eds.), Theory of the EEG Inverse Problem. Artech House,
Boston, Ch., pp. 121–140 (Chapter 5).
Pascual-Marqui, R.D., Lehmann, D., Koukkou, M., Kochi, K., Anderer, P., Saletu, B., Tanaka,
H., Hirata, K., John, E.R., Prichep, L., Biscay-Lirio, R., Kinoshita, T., 2011. Assessing interactions in the brain with exact low-resolution electromagnetic tomography. Philos.
Trans. R. Soc. A Math. Phys. Eng. Sci. 369 (1952), 3768–3784 (URL http://rsta.
royalsocietypublishing.org/content/369/1952/3768.abstract).
Pereda, E., Quiroga, R.Q., Bhattacharya, J., 2005. Nonlinear multivariate analysis of
neurophysiological signals. Prog. Neurobiol. 77 (1–2), 1–37 (URL http://www.
sciencedirect.com/science/article/pii/S030100820500119X).
Schoffelen, J.-M., Gross, J., 2009. Source connectivity analysis with MEG and EEG. Hum.
Brain Mapp. 30 (6), 1857–1865. http://dx.doi.org/10.1002/hbm.20745.
Schreiber, T., 2000. Measuring information transfer. Phys. Rev. Lett. 85, 461–464. http://
dx.doi.org/10.1103/PhysRevLett.85.461 (Jul).
Schwilden, H., 2006. Concepts of EEG processing: from power spectrum to bispectrum,
fractals, entropies and all that. Best Pract. Res. Clin. Anaesthesiol. 20 (1), 31–48
(URL http://www.sciencedirect.com/science/article/pii/S1521689605000728).
Sekihara, K., Owen, J., Trisno, S., Nagarajan, S., 2011. Removal of spurious coherence in
MEG source–space coherence analysis. IEEE Trans. Biomed. Eng. 58 (11),
3121–3129.
Shahbazi Avarvand, F., Ewald, A., Nolte, G., 2012. Localizing true brain interactions from
EEG and MEG data with subspace methods and modified beamformers. Comput.
Math. Methods Med. http://dx.doi.org/10.1155/2012/402341.
161
Shils, J.L., Litt, M., Skolnick, B.E., Stecker, M.M., 1996. Bispectral analysis of visual interactions in humans. Electroencephalogr. Clin. Neurophysiol. 98 (2), 113–125 (URL
http://www.sciencedirect.com/science/article/pii/0013469495002308).
Srinivasan, R., Winter, W.R., Ding, J., Nunez, P.L., 2007. EEG and MEG coherence: measures
of functional connectivity at distinct spatial scales of neocortical dynamics.
J. Neurosci. Methods 166 (1), 41–52 (URL http://www.sciencedirect.com/science/
article/pii/S01650270070030).
Stinstra, J.G., Peters, M.J., 1998. The volume conductor may act as a temporal filter on the
ECG and EEG. Med. Biol. Eng. Comput. 36 (6), 711–716. http://dx.doi.org/10.1007/
BF02518873.
Tallon-Baudry, C., Bertrand, O., Delpuech, C., Pernier, J., 1996. Stimulus specificity of phaselocked and non-phase-locked 40 Hz visual responses in human. J. Neurosci. 16 (13),
4240–4249 (URL http://www.jneurosci.org/content/16/13/4240.abstract).
Van Ness, J.W., 1966. Asymptotic normality of bispectral estimates. Ann. Math. Stat. 37
(5), 1257–1272 (URL http://www.jstor.org/stable/2239079).
Villa, A.E.P., Tetko, I.V., 2010. Cross-frequency coupling in mesiotemporal EEG recordings
of epileptic patients. J. Physiol. Paris 104 (3–4), 197–202 (URL http://www.
sciencedirect.com/science/article/pii/S0928425709001065).
Vinck, M., Oostenveld, R., van Wingerden, M., Battaglia, F., Pennartz, C.M.A., 2011.
An improved index of phase-synchronization for electrophysiological data in the
presence of volume-conduction, noise and sample-size bias. NeuroImage 55 (4),
1548–1565 (URL http://www.sciencedirect.com/science/article/pii/S10538119110009).
Wang, X., Chen, Y., Ding, M., 2007. Testing for statistical significance in bispectra: a surrogate data approach and application to neuroscience. IEEE Trans. Biomed. Eng. 54 (11),
1974–1982.
Winter, W.R., Nunez, P.L., Ding, J., Srinivasan, R., 2007. Comparison of the effect of volume
conduction on EEG coherence with the effect of field spread on MEG coherence. Stat.
Med. 26 (21), 3946–3957. http://dx.doi.org/10.1002/sim.2978.
Womelsdorf, T., Fries, P., 2006. Neuronal coherence during selective attentional processing and sensory–motor integration. J. Physiol. Paris 100 (4), 182–193 (URL http://
www.sciencedirect.com/science/article/pii/S092842570700006X).
Yao, D., 2001. A method to standardize a reference of scalp EEG recordings to a point at infinity. Physiol. Meas. 22 (4), 693 (URL http://stacks.iop.org/0967-3334/22/i=4/a=305).
Yao, D., Wang, L., Oostenveld, R., Nielsen, K.D., Arendt-Nielsen, L., Chen, A.C.N., 2005. A
comparative study of different references for EEG spectral mapping: the issue of
the neutral reference and the use of the infinity reference. Physiol. Meas. 26 (3),
173 (URL http://stacks.iop.org/0967-3334/26/i=3/a=003).